Logistic regression
Predict a yes-or-no outcome and read the result as an odds ratio.
When to use it
Use logistic regression when the outcome is two categories, not a score. For example: does more practice make a pass more likely? The slope is on the log-odds scale. The odds ratio is the easier number: how many times larger the odds of passing are for one extra unit of the predictor.
Assumptions
The outcome is binary. The log-odds of the outcome are a straight line in the predictors. Rows are independent. A predictor that almost perfectly separates pass from fail makes the odds ratio and its interval unusable. Tensr does not run a separation check for you.
Running it in Tensr
Options
Prop
Type
The request also accepts classification_cutoff (0.5), max_iter (20), include_constant (true), and hosmer_lemeshow (false). The analyze dispatcher does not pass those four through, so they do not change the fit. cluster_by, reference_levels, interactions, scales, predict_at, others_at, and derive_hours_since_group_min are forwarded for advanced runs.
Reading the output
Whether the student passed, predicted from study hours. The pass rate sits near the middle, and hours shifts it rather than separating the two groups.
| Term | Estimate | Odds ratio | 95% CI (OR) | p-value |
|---|---|---|---|---|
| Intercept | -1.531 | 0.216 | [0.037, 1.258] | .088 |
| hours | 0.299 | 1.348 | [0.956, 1.901] | .088 |
The primary result is not significant (p = .088). This is not large enough to treat the comparison this page is about as a reliable association. The result is non-significant: the data are still compatible with no effect. Metrics: Pseudo R² = 0.023; Observations = 96.
Reporting (APA 7)
This result is not significant (p = .088). Report the estimate with that p, and do not describe the pattern as a reliable effect.
Coming from SPSS
Analyze → Regression → Binary Logistic.
SPSS prints a classification table and can print Hosmer–Lemeshow from Options. Tensr’s report for this run printed coefficients, odds ratios, and pseudo R².
Related
A straight-line model for a numeric score is linear regression. The same binary outcome with a probit link is probit regression.